Model discussion - hunt for cold - Christmas countdown

[iowpompeylee]

I’ve just had a Twitter chat with a meteorologist and he expects everywhere in the country to be much colder by Boxing Day.still time for change Xmas day but the far south chances are reducing can’t call it yet another 24-48 hours for the details! Just shows how crazy this situation is! 

[mulzy]

3 minutes ago, Kasim Awan said:

What is the clustering between ecm op & gfs op solution within the eps?

We need to wait for the clusters to come out, but the mean eps by day 7 is quite a bit colder than last night's suite.  I have a feeling the weighting of the ecm det solution is much lower in this morning's suite when compared to last night.

Looks like the eps is following the gefs for a change!

Edited December 21, 2021 by mulzy

[Nick2373]

38 minutes ago, frosty ground said:

Why? 
 

it’s only more credible because it’s the solution you want to see, we usually meet in the middle in these circumstances.

Spot on!!

[Lukesluckybunch]

i know everyone is keeping a lot of faith in the gfs and UKMO,especially saying the gfs is being consistent which it is.but isnt the ECM also being like that not giving into the gfs solution.i think we will sadly end up closer to the ECM maybe halfway house,the reason i am saying this as even the good looking gfs is looking quite marginal in terms of the boundy line.the latest run showed milder air in the southeast boxing day temperatures were approaching 6c!i wouldnt be surprised if the 6z is watered down to a degree.im really hoping not of course.

[Chesil View]

14 minutes ago, jellybaby1969 said:

ECM is more reliable than GFS.... And has been for many years. Meto and ECM are the models to follow.

This isn't completely true. 

Both models have done better than the other at times. In my experience, in bog standard westerly flow situations yes the ecm outperforms gfs by a small margin but in more meridional or unusual situations I put them pretty much the same for a verification. Sometimes ecm gets it other gfs gets it.

[Uncertainty]

16 minutes ago, Georgina said:

I have very very little knowledge not like you guys at all , but from watching the models for a few years it’s nearly always the gfs that comes into line with the ecm !

If we have a stubborn ecm it usually ends with the gfs downgrading and the ecm slightly moving towards the gfs to meet in the middle ! 
 

of course I’m probably wrong as I say not as experienced as you guys 

Not in the last year or so Georgina. The ECM folk helped upgrade the GFS with the advent of the FV3 core and it has trumped it’s ‘master’ on a few key occasions recently. Last two runs the ECM has slowly trended towards the UKMO/GFS/GEM solution (e.g cold air moving south through Xmas day - Boxing Day with associated snow risk). The others have broadly ‘held the line’ for the last two runs.

ECM is still doubtless the best model and I won’t fully ‘buy’ this spell until Itfalls into line however if I was to forecast it I would go with the southerly solution posited above

 

[Kasim Awan]

6 minutes ago, mulzy said:

We need to wait for the clusters to come out, but the mean eps by day 7 is quite a bit colder than last night's suite.  I have a feeling the weighting of the ecm det solution is much lower in this morning's suite when compared to last night.

Looks like the eps is following the gefs for a change!

If the weighting for the ec det is under 35% then it is very marginalised 

Gfs is more likely to verify atm

Ukv & mogreps on board

Icon & arpege not on board is also a worry

65/35

[Dave Kightley]

Mogreps for London a big upgrade with vast going for cold now... been as useful as any model so can’t be that great!

[Kasim Awan]

Just now, Uncertainy said:

Not in the last year or so Georgina. The ECM folk helped upgrade the GFS with the advent of the FV3 core and it has trumped it’s ‘master’ on a few key occasions recently. Last two runs the ECM has slowly trended towards the UKMO/GFS/GEM solution (e.g cold air moving south through Xmas day - Boxing Day with associated snow risk). The others have broadly ‘held the line’ for the last two runs.

ECM is still doubtless the best model and I won’t fully ‘buy’ this spell until Itfalls into line however if I was to forecast it I would go with the southerly solution posited above

 

Yes I agree with first paragraph but ecm is a hyped model

Too slow to react and too cautious most of the time

It needs a rewrite. Are you up for some intense maths

[Ali1977]

1 minute ago, Dave Kightley said:

Mogreps for London a big upgrade with vast going for cold now... been as useful as any model so can’t be that great!

Is this just out? Any images 

[Ali1977]

2 minutes ago, Ali1977 said:

Is this just out? Any images 

M4 corridor looks the hot favourite for snow on MOGREPS - but the best bit is that it backs the GFS. GFS,UKMO, MOGREPS vs ECM and ICON!! 

Edited December 21, 2021 by Ali1977

[Kasim Awan]

Mogreps here. 82% turn it cold down to the M4 corridor Xmas night.

[Georgina]

2 minutes ago, Kasim Awan said:

Yes I agree with first paragraph but ecm is a hyped model

Too slow to react and too cautious most of the time

It needs a rewrite. Are you up for some intense maths

Not at 7.24 in the morning I’m not I dislike maths at the best of times ! 
I do agree that the ecm seems very cautious but sometimes that’s not a bad thing Xx

[iowpompeylee]

Just now, Kasim Awan said:

Mogreps here. 82% turn it cold down to the M4 corridor Xmas night.

You can clearly see the cold trend increasing on these. Good signs 

[Ali1977]

4 minutes ago, Kasim Awan said:

Mogreps here. 82% turn it cold down to the M4 corridor Xmas night.

Bloody hell, serious upgrade there ️️ Get in. ECM being and outlier, I think we have the cold nailed  and if you look to the right of the graph temps are still dropping - I wonder what they are showing early Jan

Edited December 21, 2021 by Ali1977

[Radiating Dendrite]

I'm a bit confused exactly what people are now chasing?

Whilst it looks like it could turn colder, any frontal snow risk is either too far south or when further north, it is not cold enough?

[sheikhy]

Massive upgrades all round this morming!!ecm i tell ya now seems to be struggling!!it looked primed and ready to pounce when looking at the 96 hour chart but it just didnt happen!!but there are differences still especially across the pole between ecm and gfs at 120 hours!!

[knocker]

4 minutes ago, Kasim Awan said:

Yes I agree with first paragraph but ecm is a hyped model

Too slow to react and too cautious most of the time

It needs a rewrite. Are you up for some intense maths

Are you actually suggesting the ecm is a sentient being?

[supernova]

Remarkable consistency for placement of the low from MOGREPS although significant variations of its resulting exit and therefore variable wintry output. At least a couple in there like GFS control….nothing much short of epic. Several leaving most of the country practically snowless from any frontal activity too. Still plenty to figure out but overall direction of travel absolutely tantalising.

Edited December 21, 2021 by supernova

[DoctorNick]

Model latency : Specifically, latency, is the measure of time before a weather model is run and the availability of the latest data to make a forecast. For example, if a regional weather model is set to run at 12:00 (noon) for a specific location, satellite data that are received at 11:00 a.m. would be considered 1-hour latency.

[Lukesluckybunch]

3 minutes ago, Kasim Awan said:

Mogreps here. 82% turn it cold down to the M4 corridor Xmas night.

brilliant kasim 4,5,6,11,13,14 pick of the bunch here!

[Kasim Awan]

2 minutes ago, knocker said:

Are you actually suggesting the ecm is a sentient being?

Subjective observation mate

To my eye ecm just appears too clunky at picking out new trends

Could be that the model infrastructure relies heavily on a cumulative backlog of data i.e..some sort of rolling data accumulation

[SteveB]

8 minutes ago, Ali1977 said:

M4 corridor looks the hot favourite for snow on MOGREPS - but the best bit is that it backs the GFS. GFS,UKMO, MOGREPS vs ECM and ICON!! 

Me Like 

[Kasim Awan]

6 minutes ago, Georgina said:

Not at 7.24 in the morning I’m not I dislike maths at the best of times ! 
I do agree that the ecm seems very cautious but sometimes that’s not a bad thing Xx

Imo, whoever wrote the ecm does not understand the systematic flaws of using a median method when modelling a volatile pattern 

 

[SnowBlake]

Consistency doesn't make a model more credible. It could consistently show a particular outcome for a week but at T-24 change completely.

Its verification stats are a better indicator of how reliable it is. But each model seems to handle certain situations better. The thing is, we are transfixed on the NH mostly and in particular what is happening around the UK.

A 10 mile shift of a LP system 10000 miles away can cause a 20 mile shift elsewhere, which becomes a 100 mile shift 1000 miles away and so on and so forth.

Many see the 850s for example as oil in a puddle, almost 2D and forgetting that events above and below aide the changes we see. If there is a 500 mile wide area with a 1 or 2c lesser temperature difference, this can cause all sorts of havoc. 1 or 2c can be the difference between a frozen lake, cloud build up, evaporative cooling... tshirt or jumper.

No model ever has been 100% at T-24, but can give us a very good idea of what will happen.

T-48-96, again, an indication.

T-120, an indication perhaps but I wouldn't plan a day around this.

Beyond that, we have seen too significant a change many a time to worry about it.

We know that events such as SSW can bring us colder weather a couple of weeks down the line, but not guaranteed.

Every model this past week has toyed with an idea of colder weather, some with dumpings of snow, some cold and wet, some cold and dry, some with average temperatures, and 2 days later showing a different one of the aforementioned, but for me, I have no preference over any model, and I certainly wouldn't pay for premium viewing, nor for any non-weather model... *cough*

Meanwhile here in Eastern Latvia, minus 12.8 and I brought button up fly jeans on this trip. Could do with having zips today outside. Will have to dig out my thermal undies.

Čau for now.